Time Series Data- Fish Diet

My time series spans 13 years. There are 30-60 fish stomachs for each year of the study. Each stomach has typically 3-8 prey items inside. The prey items within the stomachs are the point of interest in this study.

I have a plethora of independent variables measured from the capture of all of these fishes (i.e., depth, water temp, year, size of fish, fish weight, fish length, fish sex, etc..)

My problem is that I want to be able to predict (multiple regression?) prey fish composition or diet changes through the course of the study. With hopes of being able to predict future diet changes/inflection points.

Any help would be greatly appreciated!!